Actual Evapotranspiration Estimation Using Sentinel-1 SAR and Sentinel-3 SLSTR Data Combined with a Gradient Boosting Machine Model in Busia County, Western Kenya

Musyimi, Peter K.; Sahbeni, Ghada; Timár, Gábor; Weidinger, Tamás; Székely, Balázs

doi:10.3390/atmos13111927

Cited by 5 publications

(4 citation statements)

References 98 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…With the increase of an uneven spatial-temporal precipitation distribution in the future, short-term extreme climatic events will occur more frequently than in historical periods. The annual-scale SPEI is advantageous in assessing the changing trend and characteristics of long-term droughts; however, it is not sensitive to identifying short-term extreme climate events (Cui et al, 2021;Musyimi et al, 2023). Therefore, more meteorological drought indexes should be introduced to more accurately study and analyze the characteristics of regional long-and short-term droughts.…”

Section: Limitations and Further Researchmentioning

confidence: 99%

Prediction and assessment of meteorological drought characteristics in China based on a future climate model

Huang,

Liu,

Jia

et al. 2023

Preprint

View full text Add to dashboard Cite

Under the background of a warm and humid climate, predicting and evaluating the future pattern of wet and dry change characteristics in watersheds is of great research significance. Based on the Geophysical Fluid Dynamics Laboratory earth system coupling model and GFDL‒ESM2M climate model, which is highly applicable in China, the characteristics of standardized precipitation evapotranspiration index (SPEI) variations in China at annual and monthly scales was assessed based on downscaled climate data under four typical representative concentration pathway scenarios. The results are as follows: (1) In the future, the annual scale meteorological drought trend in China will become more "aridity", and the monthly scale dry and wet changes will be non-uniform. With the increase of carbon dioxide emission concentration, the proportion of regional aridity in China will be increased in the future. (2) The proportion of wetting area will decrease, and extreme weather events are more likely to occur. In the future, the drought duration in China will mainly be 1–4 months. (3) The return period of drought in large areas of the country will be 1–50 years, and the return period in some western regions even exceed 500 years.

show abstract

Section: Limitations and Further Researchmentioning

confidence: 99%

Prediction and assessment of meteorological drought characteristics in China based on a future climate model

Huang,

Liu,

Jia

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…However, it features a moderately coarse spatial resolution, ranging from 500 m to 1 km. Such temporal and spatial properties make the data sufficiently intricate to be incorporated into various applications, encompassing factors such as soil moisture content, surface heterogeneity, and vegetation cover [52][53][54]. In the present study, we acquired the Sentinel-3 SLSTR Level 2 product sensed on 26 February 2021 and performed a co-registration and geometric correction.…”

Section: Remote Sensing Data and Other Featuresmentioning

confidence: 99%

“…The reliability of GBM in fitting new trees and generating accurate predictions has been well-documented across different fields. For instance, it has been successfully implemented in actual evapotranspiration estimation, allowing a better understanding of short-term drought episodes associated with surface temperature and vegetation cover changes in Kenya [52]. Further, GBM has demonstrated its effectiveness in land cover and land use assessment, enabling more precise land surface mapping and monitoring [68].…”

Section: Gradient Boosting Machine (Gbm) Versus Extreme Gradient Boos...mentioning

confidence: 99%

Crop Yield Estimation Using Sentinel-3 SLSTR, Soil Data, and Topographic Features Combined with Machine Learning Modeling: A Case Study of Nepal

Sahbeni,

Székely,

Musyimi

et al. 2023

AgriEngineering

Self Cite

View full text Add to dashboard Cite

Effective crop monitoring and accurate yield estimation are fundamental for informed decision-making in agricultural management. In this context, the present research focuses on estimating wheat yield in Nepal at the district level by combining Sentinel-3 SLSTR imagery with soil data and topographic features. Due to Nepal’s high-relief terrain, its districts exhibit diverse geographic and soil properties, leading to a wide range of yields, which poses challenges for modeling efforts. In light of this, we evaluated the performance of two machine learning algorithms, namely, the gradient boosting machine (GBM) and the extreme gradient boosting (XGBoost). The results demonstrated the superiority of the XGBoost-based model, achieving a determination coefficient (R2) of 0.89 and an RMSE of 0.3 t/ha for training, with an R2 of 0.61 and an RMSE of 0.42 t/ha for testing. The calibrated model improved the overall accuracy of yield estimates by up to 10% compared to GBM. Notably, total nitrogen content, slope, total column water vapor (TCWV), organic matter, and fractional vegetation cover (FVC) significantly influenced the predicted values. This study highlights the effectiveness of combining multi-source data and Sentinel-3 SLSTR, particularly proposing XGBoost as an alternative tool for accurately estimating yield at lower costs. Consequently, the findings suggest comprehensive and robust estimation models for spatially explicit yield forecasting and near-future yield projection using satellite data acquired two months before harvest. Future work can focus on assessing the suitability of agronomic practices in the region, thereby contributing to the early detection of yield anomalies and ensuring food security at the national level.

show abstract

“…Esses sensores operam na banda C (C-band) com comprimento de onda entre 3.75 cm e 7.5 cm, o que permite uma menor interação com as partículas presentes nas nuvens e facilita a penetração do sinal. Alguns estudos estão sendo desenvolvidos para utilizar dados provenientes desses sensores na estimativa da evapotranspiração (KAPLAN et al, 2021(KAPLAN et al, , 2023MUSYIMI et al, 2022;OUAADI et al, 2023). Esses estudos estão direcionados para a quantificação do coeficiente de cultivo, representando um avanço, pois eliminaria a dependência de coeficientes tabelados utilizados no manejo da irrigação.…”

Section: Conclusãounclassified

Modelagem da evapotranspiração da cana-de-açúcar utilizando abordagens de sensoriamento remoto e machine learning

Santos

View full text Add to dashboard Cite

O uso de sensoriamento remoto combinando informações de estações meteorológicas tem sido estudado como uma alternativa a estimativa da demanda hídrica, porém algumas limitações ainda persistem, requerendo novas metodologias que possa superá-las. Desta forma, essa tese teve como objetivo geral estimar a ETa (evapotranspiração atual) da cana-de-açúcar utilizando machine learning com base em dados obtidos por sensoriamento remoto ativo e passivo, bem como múltiplas fontes de informação. Os objetivos específicos foram divididos em três capítulos: 1) Estimar a evapotranspiração utilizando machine learning e sensoriamento remoto passivo; 2) Estimar a evapotranspiração utilizando machine learning e sensoriamento remoto passivo sem a necessidade de estação meteorológica em campo; e 3) Estimar a evapotranspiração utilizando machine learning e sensoriamento remoto ativo na presença de nuvens. Para tal, foram adquiridos dados dos sensores OLI (Operational Land Imager) e TIRS (Thermal Infrared Sensor) embarcados nos satélites Landsat 8 e 9, a fim de determinar a ET rF (fração evapotranspirativa) e posterior ETa utilizando o modelo METRIC (Mapping Evapotranspiration at High Resolution with Internalized Calibration). Essas variáveis foram utilizadas como resposta, enquanto as variáveis explicativas foram obtidas dos sensores MSI (Multispectral Instrument), embarcado no Sentinel-2, e SAR (Synthetic Aperture Radar), no Sentinel-1. Os dados meteorológicos, climatológicos, de solos, elevação do terreno e astronômicas, foram utilizados para garantir um treinamento e desempenho adequados dos algoritmos de machine learning. No primeiro capítulo, foi possível observar a capacidade dos modelos de machine learning em estimar a ETrF utilizando dados do sensor MSI, que não possui banda termal. O modelo que obteve melhor desempenho nessa abordagem foi o XgbLinear, com R² = 0.91 e RMSE = 0.10, apresentando resultados satisfatórios. Os valores estimados de ETrF, neste capítulo, quando multiplicados pela ET r (Evapotranspiração de referência alfafa), forneceram a ETa, que é utilizada para determinar a reposição adequada da irrigação. Os resultados do segundo capítulo mostraram que os modelos de machine learning, utilizando dados de sensoriamento remoto passivo sem banda termal em conjunto com outras variáveis explicativas, permitiu quantificar diretamente a ET a, sem a necessidade do valor de ETr determinado via estação meteorológica. Dos 13 modelos utilizados neste capítulo, o brnn (Bayesian Regularized Neural Networks) se destacou, com R² de 0.73 e RMSE de 1.10, respectivamente. Por fim, os resultados do terceiro capítulo foram promissores, pois foi possível modelar e quantificar a ETa na cana-de-açúcar mesmo na presença de nuvens. O brnn novamente se destacou, apresentando R² de 0.45 e RMSE de 1.5. Apesar das métricas inferiores, devido às limitações espectrais do sensor SAR, esse resultado representa um grande avanço, uma vez que o sensor SAR pode ser utilizado quando não é possível o uso de sensores remotos passivos. Com base em todos esses resultados, é possível inferir que os algoritmos de machine learning XgbLinear e brnn, devidamente treinados com dados obtidos por sensoriamento remoto passivo e ativo, juntamente com outras fontes de informação, podem ser eficazes na estimativa da evapotranspiração atual da cana-de-açúcar. Palavras-chave: Satélite. Machine learning. Agricultura irrigada.

show abstract

Actual Evapotranspiration Estimation Using Sentinel-1 SAR and Sentinel-3 SLSTR Data Combined with a Gradient Boosting Machine Model in Busia County, Western Kenya

Cited by 5 publications

References 98 publications

Prediction and assessment of meteorological drought characteristics in China based on a future climate model

Prediction and assessment of meteorological drought characteristics in China based on a future climate model

Crop Yield Estimation Using Sentinel-3 SLSTR, Soil Data, and Topographic Features Combined with Machine Learning Modeling: A Case Study of Nepal

Modelagem da evapotranspiração da cana-de-açúcar utilizando abordagens de sensoriamento remoto e machine learning

Contact Info

Product

Resources

About